Conformational Control in the Rebeccamycin Class of Indolocarbazole Glycosides

Article abstract of DOI:10.1021/jo990296v

Indolocarbazole glycosides are balanced between two conformations: a "closed" conformation containing a cyclic hydrogen bond between the indolocarbazole NH and the pyranose oxygen and an "open" conformation in which the indolocarbazole NH is hydrogen bond

Full text of DOI:10.1021/jo990296v

5670
J . Org. Chem. 1999, 64, 5670-5676
Con for m a tion a l Con tr ol in th e Rebecca m ycin Cla ss of
In d oloca r ba zole Glycosid es
Eric J . Gilbert, J ohn D. Chisholm, and David L. Van Vranken*
Department of Chemistry, The University of California, Irvine, California 92697-2025
Received February 17, 1999
Indolocarbazole glycosides are balanced between two conformations: a “closed” conformation
containing a cyclic hydrogen bond between the indolocarbazole NH and the pyranose oxygen and
an “open” conformation in which the indolocarbazole NH is hydrogen bonded to solvent. The open
conformation never has a commanding advantage, even in DMSO, but in nonpolar environments
the cyclic conformation predominates.
Sch em e 1
In tr od u ction
The indolocarbazole glycosides have two basic confor-
mational designs. The staurosporine class is rigid because
the pseudocarbohydrate moiety is connected to the in-
dolocarbazole in two places. In contrast, the rebeccamycin
class seems to offer greater flexibility because it is
connected only at the anomeric position as a â-glycoside.
1
For solubility, H NMR data for most of the indolocar-
bazole glycosides has been obtained in highly polar
solvents where two conformations are observed. This
conformational dichotomy is widely recognized yet un-
explained. In this paper we show that the indolocarbazole
NH forms a cyclic hydrogen bond with the pyranose
oxygen-an effect that is only apparent in nonpolar
environments. This conformational control element may
preorganize the molecule for biological activity.
Members of the rebeccamycin class of indolocarbazole
glycosides have shown activity in a variety of mouse
xenograft models and have since advanced to human
clinical trials.1-3 These antitumor natural products are
generally regarded as “inhibitors” of human topoi-
somerase I.4 Topoisomerases resolve topological problems
of double-stranded DNA by changing supercoiling topol-
ogy. Topoisomerase I relaxes supercoiling by cleaving one
strand of DNA, allowing it to unwind, and then rejoining
the temporary strand break.5 Rebeccamycins do not
inhibit the cleavage step; they inhibit the rejoining step.6
This mechanism is selective for rapidly dividing cells
which undergo programmed cell death if DNA damage
is detected.7
Resu lts a n d Discu ssion
1. Syn th esis. Mannich dimerization and glycosylation
have made it possible to easily prepare a wide range of
indolocarbazole glycosides. To understand the conforma-
tional bias of the indolocarbazole glycosides, we first
looked at a simple indolocarbazole glycoside 1a prepared
by Mannich cyclization of 1,2-bis(3-indolyl)ethane, car-
bohydrate capture of D-glucose, and oxidative aromati-
zation with DDQ. This route is general and provides
access to the tjipanazoles, which have only modest
antifungal activity, and potent antitumor antibiotics such
as AT2433-B2, a member of the rebeccamycin class of
natural products (Scheme 1).8-10
* Phone: (949) 824-5455. FAX: (949) 824-8571. E-mail: dlvanvra@
uci.edu.
(1) Ohe, Y.; Tanigawa, Y.; Fujii, H.; Ohtsu, T.; Wakita, H.; Igarashi,
T.; Minami, H.; Eguchi, K.; Shinkai, T.; Tamura, T.; Kunitoh, H.; Saijo,
N.; Okada, K.; Ogino, H.; Sasaki, Y. Abstracts: Am. Soc. Clin. Oncol.
1997, 16, Abstract 700.
(2) Eckhardt, S. G.; Sharma, S.; Kuhn, J .; Rizzo, J .; Campbell, E.;
Ho, P.; Weiss, L.; Hammond, M.; Kraynak, M.; Von Hoff, D. D.;
Rowinsky, E. K. 33rd Annual Meeting of the American Society of
Clinical Oncology, 1997, Abstract 759.
(3) Cleary, J . F.; Berlin, J . D.; Tutsch, K. D.; Arzoomanian, R. Z.;
Alberti, D.; Feierabend, C.; Simon, K.; Morgan, K.; Marnocha, R.;
Wilding, G. 33rd Annual Meeting of the American Society of Clinical
Oncology, 1997, Abstract 760.
(4) Prudhomme, M. Curr. Pharm. Design 1997, 3, 265.
(5) D’Arpa, P.; Liu, L. F. Biochim. Biophys. Acta 1989, 989, 163.
(6) Bush, J . A.; Long, B. H.; Catino, J . J .; Bradner, W. T.; Tomita,
K. J . Antibiot. 1987, 40, 668.
(7) Froelich-Ammon, S. J .; Osheroff, N. J . Biol. Chem. 1995, 270,
21429.
(8) Chisholm, J . D.; Van Vranken, D. L. J . Org. Chem. 1995, 60,
6672.
(9) Gilbert, E. J .; Ziller, J . W.; Van Vranken, D. L. Tetrahedron 1997,
53, 16553.
(10) Gilbert, E. J .; Van Vranken, D. L. J . Am. Chem. Soc. 1996, 118,
5500.
10.1021/jo990296v CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/09/1999

Conformational Control in Rebeccamycins
J . Org. Chem., Vol. 64, No. 15, 1999 5671
F igu r e 1. 1H NMR chemical shifts of the indolocarbazole NH in d6-DMSO correspond to one of three different conformations:
closed, open, or bifurcated. For compounds 3-9, the arrow indicates which NH chemical shift corresponds to the structure shown.
Sch em e 2
the indole NH underneath the plane of the pyranose in
a solvent-accessible location. In both the open and closed
conformations, the anomeric C-H bond is about 30° out
of the plane of the aromatic ring. The forces which
disfavor other conformations can be loosely described as
A1,3 strain and A1,2 strain.
The proton NMR spectra of 1b and 2b exhibit two sets
of signals in d6-DMSO, corresponding to two conforma-
tions. In one conformation, the indolocarbazole NH
appears to be bound to d6-DMSO (Scheme 2), but the NH
chemical shift of the other conformation is upfield by
about 1.5 ppm. This divisive effect of strong hydrogen-
bonding solvents on indolocarbazole glycosides is common
and also affects glycosides 3-7 in Figure 1.11-14
F igu r e 2. Molecular mechanics (MM2*/CHCl3) predicts two
low-energy conformations: closed and open. The closed con-
formation is slightly lower in energy than the open conforma-
tion.
1
2. H NMR a n d Molecu la r Mech a n ics-Tw o Con -
for m a tion s. Molecular mechanics with the MM2* force
field provides insight into the identity of the two confor-
mations. As shown in Figure 2, the closed conformation
possesses a hydrogen bond between the pyranose oxygen
and the indole NH, whereas the open conformation places
Molecular mechanics (MM2*/CHCl3) predicts that ro-
tation about the glycosidic bond will have an activation
energy of around 11 kcal/mol in glycoside 3 (Figure 2).
Qualitatively, this result calculated with chloroform
solvation is in good agreement with experiments in
DMSO. The coalescence temperature for H5′eq of 3 in d6-
DMSO (∆ν ) 140 Hz) is about 50 °C, suggesting a barrier
around 15 kcal/mol at room temperature. For glucoside
6 (Figure 1) with a disaccharide moiety, the coalescence
temperature of the anomeric proton is about 75 °C (∆ν
) 124 Hz), corresponding to a rotational barrier of about
16 kcal/mol.
(11) Chisholm, J . D.; Golik, J .; Krishnan, B.; Matson, J . A.; Van
Vranken, D. L. J . Am. Chem. Soc. 1999, 121, 3801.
(12) Bonjouklian, R.; Smitka, T. A.; Doolin, L. E.; Molloy, R. M.;
Debono, M.; Shaffer, S. A.; Moore, R. E.; Stewart, J . B.; Patterson, G.
M. L. Tetrahedron 1991, 47, 7739.
(13) Matson, J . A. U.S. Patent 4,524,145, 1985.
(14) Nettleton, D. E.; Doyle, T. W.; Krishnan, B.; Matsumoto, G. K.;
Clardy, J . Tetrahedron Lett. 1985, 26, 4011.

5672 J . Org. Chem., Vol. 64, No. 15, 1999
Gilbert et al.
Unfortunately, molecular mechanics neglects the criti-
cal role of solvents such as DMSO, which can hydrogen
bond to the indolocarbazole NH in competition with the
pyranose oxygen. The ratio of closed to open conforma
tions in different solvents decreases in the order CDCl3,
CD3NO2, d6-acetone, d6-DMSO (eq 1). As expected, this
trend follows hydrogen-bond acceptor ability better than
solvent polarity (Table 1).
3b) and the published crystal structure of rebeccamycin14
clearly demonstrate the dominance of the closed confor-
mation.
4. Con tr olled Ma n n ich Cycliza tion -A Test of
Con for m a tion a l Con tr ol. The Mannich cyclization/
glycosylation approach is powerful and allows one to
construct variants of the indolocarbazole glycosides with
surgical precision. To unambiguously assess the impor-
tance of the closed conformation, we engineered an oxa-
analogue incapable of enjoying a pyranose-indolocarba-
zole hydrogen bond.17
Regiocontrol is an essential advantage of indole-Man-
nich cyclizations. We modified the route in Scheme 1 to
include an indole-benzofuran system. Our prediction was
that protonation of the indole would lead to electrophilic
aromatic substitution of the benzofuran - not the other
way around.
(1)
The bifunctional substrate was prepared by Wittig
coupling18 of 3-formylbenzofuran19 and a skatolyl phos-
phonium ylide.20 The double bond (1:1 mixture of E and
Z isomers) was hydrogenated, and the N-tosyl group was
removed under nucleophilic conditions to afford the
Mannich cyclization substrate. The cyclization is a sat-
isfying example of regiocontrol, giving only the desired
indoline in over 90% yield. Glycosylation and rearoma-
tization gives the single atom mutant 17a .
Ta ble 1. Effect of Solven t on th e Con for m a tion of 1b
solvent
CDCl3
closed/open
ꢀ
>99:<1
90:10
5
CD3NO2
36
21
47
d6-acetone
d6-DMSO
70:30
25:75
Sch em e 3
3. Evid en ce for th e Closed Con for m a tion . To date,
all of the indolocarbazole â-pyranosides we have prepared
adopt a single conformation in nonpolar environments
such as chloroform. The NH stretch of glycosides 2b and
3 are around 3400 cm-1 in chloroform, consistent with a
hydrogen-bonded indole NH.15 In deuteriochloroform,
evidence for the closed conformation comes from steady-
state NOEs observed in glycosides such as 3 (Figure 3a).
Irradiation of the indole NH leads to enhancement of the
crossring axial protons H3′ and H5′ of the pyranose.
Irradiation of the H1′ proton of the pyranose ring leads
to enhancement of the peri proton H2 (and also H3
through excitation transfer). A similar NOE enhance-
ment was reported for the indolocarbazole glycoside RK-
286D,16 although the implication of this conformation was
not commented upon.
Two conformations are present in d6-DMSO because
there is no hydrogen bonding to stabilize either the closed
conformation or the open conformation. Analysis of
coupling constants in acetone at low temperature con-
firms that the pyranose retains the all-equatorial chair
F igu r e 3. (a) Steady-state NOEs confirm a closed conforma-
tion for 3 in deuteriochloroform. (b) Glycoside 3 crystallizes
from ethyl acetate in the closed conformation.
Additional evidence for a cyclic hydrogen bond comes
from solid-state studies. The X-ray structure of 3 (Figure
(17) Paliwel, S.; Geib, S.; Wilcox, C. S. J . Am. Chem. Soc. 1994, 116,
4497.
(18) Bellucci, G.; Chiappe, C.; Lomoro, G. Tetrahedron Lett. 1996,
37, 4225.
(19) Beckwith, A. L. J .; Meijs, G. F. J . Chem. Soc., Chem. Commun.
1981, 595.
(15) Remers, W. A. In Indoles. Part I; Houlihan, W. J ., Ed.; Wiley:
New York, 1972; p 15.
(16) Osada, H.; Satake, M.; Koshino, H.; Onose, R.; Isono, K. J .
Antibiot. 1992, 45, 278.
(20) Nagarathnam, D. J . Heterocycl. Chem. 1992, 29, 953.

Conformational Control in Rebeccamycins
J . Org. Chem., Vol. 64, No. 15, 1999 5673
form in both conformations. This situation is different
from indolocarbazole glycoside 3, in which solvent and
pyranose oxygen compete for the indolocarbazole NH.
Once 17a is acetylated, the importance of the missing
NH is revealed. Without a cyclic hydrogen bond to fix
the conformation in deuteriochloroform, the dramatic
distinction between glycosidic rotamers is lost, and the
two conformations compete ineffectually for dominance
(Scheme 3).
5. Bifu r ca ted Hyd r ogen Bon d . Indolocarbazole â-
pyranosides with hydroxymethyl groups at the 5′-position
of the pyranose ring exist primarily in one conformation,
even in d6-DMSO. The 6′-OH group of glycosides 1a , 2a ,
and 8-10 offers the opportunity for a bifurcated hydro-
gen bond that exerts a pincerlike grip on the indole NH.
For such derivatives, the major bifurcated conformation
(>95%) is generally accompanied by a much smaller
amount (<5%) of the open conformation. Both conforma-
tions exhibit downfield NH shifts above 11 ppm, consis-
tent with strong hydrogen bonding.
1
F igu r e 5. Downfield H NMR shifts in d6-DMSO suggest that
electron-withdrawing groups enhance the acidity of indole
protons. In indolocarbazole glycoside natural products, aryl
substituents are found in positions that acidify the free NH.
fying effect improves hydrogen bonding to acceptor
solvents and to the pyranose oxygen, but in the absence
of a hydrogen bond acceptor solvent, only the pyranose
oxygen can benefit from this effect.
The chloro substituent of AT-2433 A1/A2 appears to
be too far from the free NH to affect the acidity. Molecular
mechanics (MM2*/CDCl3) suggests that this pattern of
chloro substitution destabilizes the open conformation by
about 1 kcal/mol. Thus, the chloro groups of rebeccamycin
favor the closed conformation in two different ways. The
chloro group peri to the free NH enhances its ability to
act as a hydrogen bond donor, whereas the chloro group
peri to the N-glycoside destabilizes the open conforma-
tion.
F igu r e 4. Bifurcated conformation in glycoside 1a in d6-
DMSO and an isonucleoside 18.
Evidence for the bifurcated conformation comes from
NOE studies of glucoside 1a in d6-DMSO. The small J 5′,6′
coupling constants suggest that the hydroxy group is
pointed up (antiperiplanar to the 5′-H), and this is
confirmed by a 6% steady-state NOE between the 6′-OH
and the indolocarbazole NH. A similar bifurcated hydro-
gen bond has been observed in the crystal structure of a
ribofuranoside derivative 1821 (Figure 4). As expected, the
open conformation of the indolocarbazole glycosides reas-
serts itself when the 6-hydroxy group is acetylated.
The bifurcated arrangement is probably unnecessary
for biological activity. Glycosylation of the 6′-OH in
AT2433-B1 (7) and AT2433-A1 (12) prevents participa-
tion in a bifurcated hydrogen bond, an effect attributable
to sterics. Yet the AT2433 antibiotics are potent antitu-
mor agents. Rebeccamycin (11) also shows no evidence
for a bifurcated hydrogen bond in d6-DMSO. Rather, the
chemical shift of the indole NH supports a closed con-
formation that involves only the pyranose oxygen. In this
case, the chloro group peri to the free NH discourages
participation by the 6′-OH.
Con clu sion
Indolocarbazole glycosides are balanced between two
conformations: a closed conformation containing a cyclic
hydrogen bond and an open conformation that is favored
by intermolecular hydrogen bonding. The open conforma-
tion never has a commanding advantage, even in DMSO.
In addition, one finds that indolocarbazole glycoside
natural products are substituted in such a way as to
strengthen preorganization in the closed conformation.
In fact, there is no evidence that either rebeccamycin (11)
or AT-2433-A1 (12) ever adopt an open conformation,
even in DMSO.
The forces that favor the closed conformation are
intrinsic to indolocarbazole glycoside structure. Conse-
quently, any complex involving an indolocarbazole gly-
coside must act either with or against these forces.
Currently, there are two basic models for the ternary
interaction of topoisomerase I and DNA with inhibitors:
the intercalation model depicted by Prudhomme and co-
workers for indolocarbazole glycosides26 and the base-
flipped shelf model developed by Hol and co-workers for
camptothecin.27 There is still no detailed mechanistic
model for the interaction of indolocarbazole glycosides
6. Aglycon e Su bstitu en t Effects. In the indolocar-
bazole glycoside natural products, substituents are found
where they can best improve the hydrogen bond donor
ability of the indolocarbazole NH. In d6-DMSO such
substitutents lead to downfield shifts of the indolocar-
bazole NH. The electronic pathway is traced in Figure 5
for several aglycones of natural products.22-25 This acidi-
(21) Worthington, V. L.; Fraser, W.; Schwalbe, C. H. Carbohydr. Res.
1995, 275, 275.
(22) Fabre, S.; Prudhomme, M.; Sancelme, M.; Rapp, M. Bioorg.
Med. Chem. 1994, 2, 73.
(23) Hughes, I.; Nolan, W. P.; Raphael, R. A. J . Chem. Soc., Perkin
Trans. 1 1990, 2475.
(24) Fabre, S.; Prudhomme, M.; Rapp, M. Bioorg. Med. Chem. Lett.
1992, 2, 449.
(25) Bergman, J .; Pelcman, B. J . Org. Chem. 1989, 54, 824.
(26) Bailly, C.; Riou, J . F.; Colson, P.; Houssier, C.; Rodrigues-
Pereira, E.; Prudhomme, M. Biochemistry 1997, 36, 3917.

5674 J . Org. Chem., Vol. 64, No. 15, 1999
Gilbert et al.
chromatography (40% EtOAc/Hex) afforded (()-6-methyl-4b,-
4c,6,7,7a,12,12b,13-octahydroindolo[2,3-a]pyrrolo[3,4-c]carba-
zole-5,7-dione (0.760 g, 91% yield) as an orange foam: mp 178-
with DNA and topoisomerase I. Because the structure
of the covalent complex between DNA and human topo-
isomerase has recently been solved, such a model for
indolocarbazole glycoside inhibition is certainly within
reach.
182 °C; Rf ) 0.36 (40% EtOAc/Hex); IR (film) 3368, 1607 cm-1
;
1H NMR (500 MHz, CD3CN, 500 MHz) δ 9.19 (s, 1H), 7.82 (d,
J ) 7.6 Hz, 1H), 7.30 (d, J ) 8.0 Hz, 1H), 7.16 (d, J ) 7.6,
1H), 7.10 (t, J ) 7.6, 1H), 7.02 (t, J ) 7.6, 1H), 6.94 (t, J )
7.6, 1H), 6.67 (t, J ) 7.6, 1H), 6.56 (d, J ) 7.6, 1H), 4.97 (s,
1H), 4.80 (d, J ) 7.6, 1H), 4.30 (d, J ) 7.2, 1H), 4.20 (d, J )
7.6, 1H), 4.01 (dd, J ) 7.6, 2.0, 1H), 2.82 (s, 3H); 13C NMR
(125 MHz, CDCl3) δ 179.3, 177.9, 150.6, 137.4, 135.2, 130.1,
129.0, 126.8, 123.6, 123.3, 121.4, 120.4, 119.6, 111.9, 110.8,
106.4, 54.5, 41.1, 40.7, 38.8, 25.1; MS (CI) 345, 344, 343, 285,
232; HRMS (CI) calcd for C21H17N3O2 343.1321, found 343.1320.
Anal. Calcd for C21H17N3O2: C, 73.45; H, 4.99; N, 12.24.
Found: C, 73.51; H, 5.04; N, 12.27.
Exp er im en ta l Section
11-(â-D-Glu cop yr a n osyl)-in d olo[2,3-a ]ca r ba zole (1a ).
To (()-5,6,6a(s),11a(s)-tetrahydroindolo[2,3-a]carbazole (0.40
g, 1.54 mmol) suspended in MeOH (10 mL) was added
D-glucose (0.83 g, 4.6 mmol). The mixture was stirred at reflux
for 24 h and then concentrated in vacuo. The residue was
preadsorbed on silica gel, and the glycosylation products were
separated from the excess D-glucose by silica gel chromatog-
raphy (5% MeOH/EtOAc) to afford a 1:1 mixture of diastere-
omers (0.51 g, 78%). The mixture of diastereomers was carried
on directly by taking up into 1,4-dioxane (4.8 mL). DDQ (0.56
g, 2.05 mmol) was added, and the reaction mixture was stirred
at room temperature for 12 h. Saturated NaHCO3 and EtOAc
were added, and the solution was stirred vigorously for 15 min.
The mixture was then extracted with EtOAc. The organic
layers were combined, washed with H2O and brine, and dried
over MgSO4. Filtration and evaporation of the solvent in vacuo
afforded a residue that was preadsorbed on silica gel. Purifica-
tion by silica gel chromatography (5% MeOH/EtOAc) provided
(()-6-Methyl-4b,4c,6,7,7a,12,12b,13-octahydroindolo[2,3-a]-
pyrrolo[3,4-c]carbazole-5,7-dione (101 mg, 0.292 mmol) was
dissolved in 3 mL of methanol. R-D-Glucose (158 mg, 0.876
mmol) was then added, and the reaction mixture was heated
to reflux. After 60 h the reaction mixture was allowed to cool
to room temperature and purified by silica gel chromatography
(9% MeOH/CHCl3) to give a 1:1 mixture of diastereomers by
1H NMR (0.128 g, 87%). This was dissolved in 2 mL of 1,4-
dioxane, and DDQ (127 mg, 0.56 mmol) was added. After 24 h
the reaction mixture was taken up in 40 mL of ethyl acetate
and washed with saturated NaHCO3 solution. The organic
layer was then dried (Na2SO4) and concentrated. Purification
using silica gel chromatography (3% MeOH/EtOAc) gave 2a
1a (0.37 g, 73%): [R]23 ) -30.0° (c 0.5, 10% MeOH/CHCl3);
D
mp 196-200 (dec) °C (CH2Cl2); Rf ) 0.40 in 5% MeOH/EtOAc;
1
IR (KBr) 3571, 3058, 2923 cm-1; H NMR (500 MHz, DMSO-
(107 mg, 84%) as an orange solid: mp 331 °C (dec); [R]23
)
D
d6) δ 11.04 (s, 1Η), 8.17 (d, J ) 7.9, 1H), 8.15 (d, J ) 8.0, 1H),
8.02 (d, J ) 8.2, 1H), 7.97 (d, J ) 8.2, 1H), 7.81 (d, J ) 8.4,
1H), 7.55 (d, J ) 8.0, 1H), 7.39 (m, 2H), 7.21 (m, 2H), 6.10 (d,
J ) 8.4, 1H), 5.90 (s (br), 1H), 5.32 (s (br), 1H), 5.10 (s (br),
1H), 4.84 (s (br), 1H), 4.06 (m, 1H), 4.01 (m, 1H), 3.87 (d, J )
9.8, 1H), 3.79 (d, J ) 11.0, 1H), 3.56 (m, 2H); 13C NMR (125
MHz, DMSO-d6) δ 140.5, 139.5, 126.4, 124.8, 124.58, 124.55,
123.8, 123.4, 122.2, 121.4, 119.7, 119.4, 119.3, 119.0, 112.9,
111.7, 111.5, 111.3, 84.3, 78.4, 76.9, 73.3, 67.6, 58.5; LRMS
(FAB) 418 (100); HRMS (CI) calcd for C24H22N2O5 418.1528,
found 343.1527.
+198.5 (c 0.9, THF); Rf ) 0.38 (5% MeOH/EtOAc); IR (film)
3341, 2990, 1688 cm-1; 1H NMR (500 MHz, acetone-d6) δ 11.30
(s, 1H), 9.08 (d, J ) 7.6, 1H), 8.84 (d, J ) 8.0, 1H), 7.79 (d, J
) 8.3, 1H), 7.51 (d, J ) 8.3, 1H), 7.48 (t, J ) 7.7, 1H), 7.36 (t,
J ) 8.2 Hz, 1H), 7.27 (t, J ) 7.6, 1H), 7.22 (t, J ) 7.6, 1H),
6.30 (d, J ) 9.2, 1H), 5.11 (t, J ) 3.7, 1H), 4.77 (s, 1H), 4.66 (s,
1H), 4.48 (d, J ) 5.2, 1H), 4.38 (dd, J ) 9.6, 3.6, 1H), 4.08-
4.11 (m, 3H), 3.90-3.94 (m, 2H), 2.79 (s, 3H); 13C NMR (125
MHz, acetone-d6) δ 170.4, 170.3, 143.2, 142.1, 130.8, 129.2,
127.6, 125.8, 125.4, 122.7, 122.6, 121.4, 121.1,121.0, 119.7,
119.1, 118.7, 112.6, 112.5, 111.8, 86.1, 80.1, 78.2, 74.6, 69.0,
60.1, 23.5; MS (FAB) 517, 502, 501, 368, 339, 273, 255; HRMS
(FAB) calcd for C27H23N3O7 501.1536, found 501.1522. Anal.
Calcd for C27H23N3O7‚H2O: C, 62.45; H, 4.85; N, 8.09. Found:
C, 62.46; H, 4.84; N, 8.03.
11-(â-D-Tet r a -O-a cet ylglu cop yr a n osyl)-in d olo[2,3-a ]-
ca r ba zole (1b). To 1a (0.30 g, 0.07 mmol) in pyridine (0.3
mL) was added acetic anhydride (0.059 g, 0.57 mmol). The
reaction mixture was stirred at room temperature for 24 h and
then concentrated in vacuo. The residue was preadsorbed on
silica gel and purified by silica gel chromatography (30%
EtOAc/Hex) to afford 1b (0.040 g, 98%): [R]23D ) +23.4° (c 0.5,
10% MeOH/CHCl3); mp 296-298 °C (CH2Cl2); Rf ) 0.29 in 30%
6-Meth yl-12-[2,3,4,6-tetr a-O-acetyl-â-D-glu copyr an osyl]-
6,7,12,13-tetr a h yd r oin d olo[2,3-a ]p yr r olo[3,4-c]ca r ba zole-
5,7-d ion e (2b). Compound 2a (0.06 mmol, 29 mg) was
dissolved in 1 mL of pyridine. Acetic anhydride (1.35 mmol,
100 µL) was then added. After 25 h the reaction was taken up
in ethyl acetate and washed with 1 N HCl. The organic layer
was then dried (Na2SO4) and concentrated. Purification by
silica gel chromatography (60% EtOAc/Hex) afforded 2b (33
1
EtOAc/Hex; IR (KBr) 3418, 3048, 2948, 1750 cm-1; H NMR
(400 MHz, CDCl3) δ 9.37 (s, 1H), 8.17 (d, J ) 7.8, 1H), 8.09 (d,
J ) 7.7, 1H), 8.06 (d, J ) 8.2, 1H), 7.92 (d, J ) 8.2, 1H), 7.66
(d, J ) 8.0, 1H), 7.47 (m, 3H), 7.31 (d, J ) 7.9, 1H), 7.27 (d, J
) 8.0, 1H), 6.09 (d, J ) 8.8, 1H), 5.72 (t, J ) 9.9, 1H), 5.56 (m,
2H), 4.74 (d, J ) 10.3, 1H), 4.33 (m, 2H), 2.27 (s, 3H), 2.14 (s,
3H), 1.94 (s, 3H), 1.13 (s, 3H); a carbon resonance in the 109-
130 ppm range could not be resolved; 13C NMR (125 MHz,
CDCl3) δ 170.5, 170.2, 169.3, 167.7, 139.9, 139.7, 126.6, 125.12,
125.1, 124.9, 123.39, 123.35, 120.8, 120.2, 120.1, 120.0, 114.3,
112.1, 111.6, 109.7, 84.4, 75.7, 73.1, 71.2, 67.3, 61.2, 21.2, 20.6,
20.5, 19.2, 18.6; LRMS (FAB) 586 (100); HRMS (CI) calcd for
mg, 83%): mp 268-270 °C (ethyl acetate); [R]23 ) +67.4 (c
D
1.65, CHCl3); Rf ) 0.49 (50% EtOAc/Hex); IR (CHCl3) 3393,
3024, 2942, 1753, 1698, 1582 cm-1; 1H NMR (500 MHz, CDCl3)
δ 9.90 (s, 1H), 9.26 (d, J ) 8.0, 1H), 9.22 (d, J ) 8.0, 1H), 7.66
(d, J ) 8.0, 1H), 7.56-7.59 (m, 2H), 7.51 (d, J ) 8.4, 1H), 7.41
(dt, J ) 7.1, 7.5, 2H), 6.10 (d, J ) 9.2, 1H), 5.68 (dd, J ) 9.9,
9.6, 1H), 5.58 (dd, J ) 9.5, 9.2, 1H), 5.47 (dd, J ) 9.5, 9.2,
1H), 4.81 (dd, J ) 12.7, 3.2, 1H), 4.47-4.42 (m, 2H), 3.24 (s,
3H), 2.29 (s, 3H), 2.14 (s, 3H), 1.92 (s, 3H), 1.07 (s, 3H); 13C
NMR (125 MHz, CDCl3) δ 170.5, 169.9, 169.8, 169.7, 169.2,
167.8, 141.1, 140.7, 129.9, 128.3, 127.4, 127.2, 126.0, 125.8,
122.7, 122.6, 122.0, 121.8, 121.4, 119.8, 119.4, 119.2, 111.4,
109.5, 84.3, 76.2, 70.7, 67.1, 61.0, 23.7, 21.2, 20.5, 20.4, 19.0;
MS (FAB+) 669, 368, 339, 282; HRMS (FAB+) calcd for
C
32H30N2O9 586.1951, found 586.1951.
6-Meth yl-12-[â-D-glu copyr an osyl]-6,7,12,13-tetr ah ydr oin -
d olo[2,3-a ]p yr r olo[3,4-c]ca r ba zole-5,7-d ion e (2a ). cis-N-
Methyl-2,3-bis(3-indolyl) succinimide (0.836 g, 2.43 mmol) was
dissolved in 20 mL of trifluoroacetic acid. After 30 min, the
TFA was removed in vacuo, and the reaction mixture was
taken up in 50 mL of ethyl acetate. The reaction mixture was
then poured into saturated aqueous NaHCO3. The layers were
separated, and the aqueous layer was extracted with ethyl
acetate. The combined organic layers were then dried (Na2-
SO4) and concentrated in vacuo. Purification by silica gel
C
C
35H31N3O11 669.1958, found 669.1957. Anal. Calcd for
35H31N3O11: C, 62.78; H, 4.67; N, 6.27. Found: C, 62.63; H,
4.69; N, 6.24.
11-(Tr i-O-a cetyl-â-D-xylop yr a n osyl)in d olo[2,3-a ]ca r ba -
zole (3). To a stirred suspension of (()-5,6,6a(s),11a(s)-
tetrahydroindolo[2,3-a]carbazole (2.04 g, 7.84 mmol) in MeOH
(45 mL) was added D-xylose (3.54 g, 23.5 mmol). The reaction
mixture was stirred at reflux for 12 h and then concentrated
(27) Redinbo, M. R.; Stewart, L.; Kuhn, P.; Champoux, J . J .; Hol,
W. G. J . Science 1998, 1504.

Conformational Control in Rebeccamycins
J . Org. Chem., Vol. 64, No. 15, 1999 5675
in vacuo. The residue was preadsorbed on silical gel and
purified by silica gel chromatography (1% MeOH/EtOAc) to
afford a 1:1 mixture of diastereomers (2.60 g, 85%). A portion
of the mixture of diastereomers (0.21 g, 0.52 mmol) was taken
on directly by taking up into 1,4-dioxane (2.1 mL). DDQ (0.242
g, 1.06 mmol) was then added, and the reaction mixture was
stirred at room temperature for 12 h. Saturated NaHCO3 and
EtOAc were added, and the solution was stirred vigorously
for 15 min. The mixture was then extracted with EtOAc. The
organic layers were combined, washed with H2O and brine,
and dried over MgSO4. Filtration and evaporation of the
solvent in vacuo afforded a residue that was preadsorbed on
silica gel. Purification by silica gel chromatography (2% MeOH/
EtOAc) provided 11-(â-D-xylopyranosyl)indolo[2,3-a]carbazole
10.1, 1H), 7.41 (d, J ) 7.5, 1H), 7.36 (dd, J ) 7.6, 7.5, 1H),
6.09 (d, J ) 9.2, 1H), 5.60 (dd, J ) 9.0, 8.9, 1H), 5.56 (d, J )
4.0, 1H), 5.46 (t, J ) 10.1, 1H), 5.33 (t, J ) 9.2, 1H), 5.17 (t, J
) 9.9, 1H), 4.92 (dd, J ) 10.5, 3.8, 1H), 4.81 (d, J ) 4.9, 1H),
4.72 (dd, J ) 12.5, 2.4, 1H), 4.54 (t, J ) 9.3, 1H), 4.38 (td, J )
9.9, 2.4, 1H), 4.31 (dd, J ) 12.4, 3.2, 1H), 4.22 (dd, J ) 12.6,
1.6, 1H), 4.04 (td, J ) 10.2, 2.5, 1H), 3.12 (s, 3H), 2.28 (s, 3H),
2.18 (s, 3H), 2.08 (s, 3H), 2.06 (s, 3H), 2.04 (s, 3H), 1.90 (s,
3H), 1.06 (s, 3H); one carbon resonance in the 110-150 ppm
range could not be resolved; 13C NMR (125 MHz, CDCl3) δ
170.5, 170.4, 169.9, 169.7, 169.5, 169.4, 169.0, 141.1, 140.6,
129.8, 128.0, 127.4, 127.1, 126.0, 125.9, 122.7, 122.3, 121.9,
121.6, 121.5, 119.7, 119.4, 119.1, 111.2, 109.5, 95.9, 83.7, 76.2,
75.3, 72.1, 71.4, 70.1, 69.1, 68.9, 67.9, 62.3, 61.3, 29.2, 23.5,
21.2, 20.7, 20.6, 20.5, 19.0. MS (FAB+) 957, 619, 307; HRMS
(FAB+) calcd for C47H47N3O19 957.2803, found 957.2824. Anal.
Calcd for C47H47N3O19: C, 58.92; H, 4.95; N, 4.39. Found: C,
59.04; H, 5.02; N, 4.30.
(0.14 g, 70%): [R]23 ) +10.4° (c 0.5, 10% MeOH/CHCl3); mp
D
288 °C (dec, CH2Cl2); Rf ) 0.55 in 1% MeOH/EtOAc; IR (KBr)
3398, 3058, 2923, cm-1; 1H NMR (300 MHz, DMSO-d6, 90 °C)
δ 10.27 (s (br), 1H), 8.12 (m, 2H), 7.95 (m, 2H), 7.77 (m, 1H),
7.69 (m, 1H), 7.38 (m, 2H), 7.21 (m, 2H), 6.02 (d, J ) 8.8, 1H),
4.84 (m, 1H), 4.65 (m (br), 2H), 4.22 (m, 1H), 3.88 (m, 2H),
3.75 (m, 1H), 3.64 (m, 1H); the 13C NMR showed two sets of
signals at room temperature which were not fully resolved in
the 110-126 ppm region and the 67-70 ppm region; 13C NMR
(125 MHz, DMSO-d6, 23 °C) δ 140.8, 139.5, 139.2, 137.6, 127.1,
126.0, 125.2, 124.9, 124.7, 124.3, 123.9, 123.4, 123.2, 122.4,
122.1, 121.6, 120.4, 119.7, 119.6, 119.4, 119.0, 114.0, 112.9,
112.6, 112.3, 111.9, 111.6, 111.4, 111.2, 87.5, 85.7, 78.0, 77.2,
73.2, 70.7, 69.6, 69.5, 69.2, 68.3; MS (FAB) 388 (100), 273 (15),
256 (39); HRMS (FAB) calcd for C23H20N2O4 388.1422, found
388.1421.
6-Meth yl-12-[O-(r-D-glu cop yr a n osyl)-(1f4)-â-D-glu co-
p yr a n osyl]-6,7,12,13-tetr a h yd r oin d olo[2,3-a ]p yr r olo[3,4-
c]ca r ba zole-5,7-d ion e (9). 6-Methyl-4b,4c,6,7,7a,12,12b,13-
octahydroindolo[2,3-a]pyrrolo[3,4-c]carbazole-5,7-dione (0.653
mmol, 224 mg), maltose (1.95 mmol, 670 mg), and (NH4)2SO4
(0.04 mmol, 1 mg) were suspended in 7 mL of methanol and
warmed to reflux. After 72 h the reaction was preabsorbed on
silica gel and purified by silica gel chromatography (20%
methanol/chloroform) to give a mixture of diastereomers (308
mg, 71%). This mixture was dissolved in 10 mL of dioxane,
and DDQ (1.02 mmol, 230 mg) was added. After 48 h the
reaction was concentrated in vacuo and purified by reverse
phase HPLC (70% methanol/30% 0.1% aqueous TFA) to give
26 mg (8%) of 6-methyl-12-[O-(R-D-glucopyranosyl)-(1f4)-â-
D-glucopyranosyl]-6,7,12,13-tetrahydroindolo[2,3-a]pyrrolo[3,4-
c]carbazole-5,7-dione (extensive decomposition was observed
with this buffer system; further experiments are in progress
to find alernative conditions for purification): mp 345 °C (dec);
To 11-(â-D-xylopyranosyl)indolo[2,3-a]carbazole (0.14 g, 0.36
mmol) in pyridine (1.2 mL) was added acetic anhydride (0.19
g, 1.82 mmol). The reaction mixture was stirred at room
temperature for 17 h and then concentrated in vacuo. The
residue was preadsorbed on silica gel and purified by silica
gel chromatography (30-45% EtOAc/Hex) to provide 3 (0.14
g, 76%): [R]23 ) -41.2° (c 1.0, CHCl3); mp 316 °C (dec,
[R]23 ) +151.0 (c 0.05, 1:1 MeOH/DMF); Rf ) 0.22 (25%
D
D
EtOAc); Rf ) 0.30 in 30% EtOAc/Hex; IR (KBr) 3402, 3040,
MeOH/CHCl3); IR (KBr) 3329, 2922, 1745, 1690, 1572 cm-1
;
2952, 1753 cm-1; H NMR (500 MHz, CDCl3) δ 9.27 (s, 1H),
1H NMR (500 MHz, DMSO-d6) δ 11.62 (s, 1H), 9.17 (d, J )
8.0, 1H), 9.10 (d, J ) 8.0, 1H), 7.96 (d, J ) 8.5, 1H), 7.75 (d, J
) 8.1, 1H), 7.55-7.61 (m, 2H), 7.36-7.40 (m, 2H), 6.36 (d, J
) 9.0, 1H), 6.04 (bs, 1H), 5.72 (bs, 1H), 5.54 (bs, 1H), 5.30 (d,
J ) 3.5, 1H), 5.10-5.21 (m, 3H), 4.69 (bs, 1H), 4.26 (bd, J )
9.6, 1H), 4.16-4.21 (m, 2H), 3.92 (t, J ) 8.4, 1H), 3.84 (bd, J
) 10.7, 1H), 3.78 (d, J ) 11.0, 1H), 3.66-3.69 (m, 1H), 3.55-
3.63 (m, 3H), 3.50 (t, J ) 9.1, 1H), 3.34-3.36 (m, 1H), 3.15 (t,
J ) 9.4, 1H); 13C NMR (DMSO-d6, 125 MHz) δ 169.7, 169.6,
142.2, 140.9, 129.5, 128.0, 127.1, 126.9, 124.3, 124.2, 121.4,
121.0, 120.7, 120.5, 120.1, 118.5, 118.4, 117.2, 112.4, 111.8,
100.9, 84.0, 77.8, 77.0, 76.0, 73.8, 73.4, 72.5, 70.0, 61.1, 58.4,
23.7. LRMS (FAB+) 686 (M + Na), 413, 365, 301; HRMS
(FAB+) calcd for C33H33N3O12 663.2064, found 663.2081.
1-(N-Tosyl-3-in d olyl)-2-(3-ben zofu r yl)eth a n e (14). To
the aldehyde 13 (0.47 g, 3.2 mmol) in methylene chloride (7
mL) at 0 °C was added 1-p-toluenesulfonyl-indole-3-methylt-
riphenylphosphonium bromide (2.24 g, 3.2 mmol), 18-crown-6
(0.05 g, 0.18 mmol), and crushed potassium hydroxide (0.39
g, 7.1 mmol). The reaction was allowed to warm to room
temperature, and stirring was contiued for 5 h. The reaction
was then added to H2O, and the mixture was extracted with
CH2Cl2. The organic layer was washed with H2O and brine
and dried over MgSO4. Filtration and evaporation of the
solvent in vacuo afforded a solid that was preadsorbed on silica
gel. Purification by silica gel chromatography (10% EtOAc/
Hex) provided a 1:1 mixture of E:Z olefin isomers (0.97 g, 73%)
that were carried on directly to the next step.
1
8.16 (d, J ) 7.8, 1H), 8.08 (d, J ) 7.8, 1H), 8.05 (d, J ) 8.1,
1H), 7.91 (d, J ) 8.1, 1H), 7.65 (d, J ) 8.0, 1H), 7.47 (m, 3H),
7.29 (m, 2H), 6.00 (d, J ) 9.0, 1H), 5.57 (t, J ) 9.5, 1H), 5.40
(m, 2H), 4.78 (m, 1H), 3.91 (t, J ) 11.2, 1H), 2.16 (s, 3H), 1.94
(s, 3H), 1.08 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 170.1, 169.9,
167.8, 140.0, 139.4, 127.5, 125.2, 125.1, 125.0, 124.9, 123.4,
123.1, 120.7, 120.1, 119.9, 114.3, 112.0, 111.5, 110.1, 110.0,
107.5, 84.9, 72.3, 71.2, 68.7, 66.2, 20.7, 20.5, 19.1; MS (FAB)
514 (100); HRMS (FAB) calcd for C29H26N2O7 514.1739, found
514.1731. Anal. Calcd for C29H26N2O7: C, 67.70; H, 5.09; N,
5.44. Found: C, 67.53; H, 5.14; N, 5.38.
6-Met h yl-12-[O-(2,3,4,6-t et r a -O-a cet yl-r-D-glu cop yr a -
n osyl)-(1f4)-2,3,6-t r i-O-a cet yl-â-D-glu cop yr a n osyl]-6,7,-
12,13-tetr a h yd r oin d olo[2,3-a ]p yr r olo[3,4-c]ca r ba zole-5,7-
dion e (4). 6-Methyl-4b,4c,6,7,7a,12,12b,13-octahydroindolo[2,3-
a]pyrrolo[3,4-c]carbazole-5,7-dione (93 mg, 0.268 mmol) and
maltose (290 mg, 0.804 mmol) were suspended in 3 mL of
methanol. (NH4)2SO4 (4 mg, 0.04 mmol) was added, and the
reaction was heated to reflux for 62 h. The reaction mixture
was then cooled to room temperature and purified by silica
gel chromatography (20% methanol/chloroform). This gave 132
mg (74% yield) of the glycosylated indolindolines, which was
dissolved in 6 mL of dioxane, and DDQ (99 mg, 0.433 mmol)
was added. The reaction was then heated to 70 °C for 16 h.
The solvent was then removed in vacuo. The crude reaction
mixture was dissolved in 5 mL of pyridine, and acetic
anhydride (27.3 mmol, 2 mL) was added. After 38 h the
reaction mixture was poured into 100 mL of 1 N HCl and
extracted with CHCl3. The extracts were then dried (Na2SO4)
and concentrated in vacuo. Purification via silica gel chroma-
tography (1:1 Hex/EtOAc) gave 4 (157 mg, 58%): mp 156-158
To the 1:1 mixture of E:Z olefin isomers (1.51 g, 3.65 mmol)
in a 250 mL Paar hydrogenation bottle was added ethyl acetate
(50 mL) followed by 10% Pd/C (226 mg, 15% w/w). The reaction
was put under a 50 psi hydrogen atmosphere on a Parr shaker
for 96 h at room temperature. The reaction mixture was then
passed through a plug of Celite, washing with chloroform. The
filtrate was concentrated in vacuo to afford a solid that was
preadsorbed on silica gel. Purification by silica gel chroma-
tography (5-10% EtOAc/Hex) provided 14 (1.16 g, 77%): mp
°C (ethyl acetate); [R]23 ) +183.7 (c 0.6, CHCl3); Rf ) 0.14
D
(50% EtOAc/Hex); IR (CHCl3) 3390, 3030, 2958, 1753, 1699,
1582 cm-1; 1H NMR (500 MHz, CDCl3) δ 9.81 (s, 1H), 9.21 (d,
J ) 7.8, 1H), 9.18 (d, J ) 7.9, 1H), 7.64 (d, J ) 8.0, 1H), 7.59
(dd, J ) 7.2, 7.8, 1H), 7.52 (dd, J ) 7.6, 7.5, 1H), 7.44 (d, J )

5676 J . Org. Chem., Vol. 64, No. 15, 1999
Gilbert et al.
128-130 °C (CH3CN); Rf ) 0.5 in 10% EtOAc/Hex; IR (KBr)
3104, 3021, 2906, 1379, 1173 cm-1; 1H NMR (500 MHz, CDCl3)
δ 7.98 (d, J ) 8.3, 1H), 7.66 (d, J ) 8.4, 2H), 7.55 (d, J ) 7.6,
1H), 7.48 (d, J ) 8.1, 2H), 7.27 (m, 6H), 7.19 (d, J ) 8.1, 2H),
3.06 (m, 4H), 2.34 (s, 3H); two carbon resonances in the 110-
150 ppm range could not be resolved; 13C NMR (125 MHz,
CDCl3) δ 155.3, 144.7, 141.3, 135.3, 135.2, 130.8, 129.8, 127.9,
126.7, 124.7, 124.3, 123.1, 122.4, 122.3, 119.4, 119.3, 113.9,
111.5, 24.6, 23.2, 21.6; MS (CI) 415 (46), 284 (100), 130 (91);
HRMS (CI) calcd for C25H21NO2 415.1242, found 415.1243.
Anal. Calcd for C25H21NO2: C, 72.27; H, 5.09; N, 3.37. Found:
C, 72.24; H, 5.11; N, 3.42.
carried on directly by taking up into 1,4-dioxane (2.5 mL). DDQ
(0.27 g, 1.17 mmol) was added, and the reaction was stirred
at room temperature for 24 h. Saturated NaHCO3 and EtOAc
were added, and the mixture was stirred vigorously for 15 min,
followed by extraction with EtOAc. The combined organic
layers were washed with H2O and brine and dried over MgSO4.
Filtration and evaporation of the solvent in vacuo afforded a
solid that was preadsorbed on silica gel. Purification by silica
gel chromatography (2% MeOH/EtOAc) provided 17a (0.16 g,
74%): [R]23 ) +112° (c 1.0, CHCl3); mp 146 °C (sintered)
D
(CHCl3); Rf ) 0.56 in EtOAc; IR (KBr) 3386, 3052, 2921, 2849
1
cm-1; H NMR (400 MHz, DMSO-d6, 120 °C) δ 8.14 (m, 3H),
1-(3-In d olyl)-2-(3-ben zofu r yl)eth a n e (15). To 14 (1.16 g,
2.80 mmol) in ethanol (22 mL) was added aqueous 3 N NaOH
(12 mL). The reaction mixture was stirred at reflux for 24 h.
Water was added (150 mL) to the cooled reaction mixture, and
the resulting white solid was filtered to provide 15 (0.67 g,
92%): mp 122-124 °C (Et2O); Rf ) 0.5 in 15% EtOAc/Hex; IR
(KBr) 3409, 3042, 2907, 2848 cm-1; 1H NMR (500 MHz, CDCl3)
δ 7.89 (s (br), 1H), 7.64 (d, J ) 7.8, 1H), 7.55 (d, J ) 7.6, 1H),
7.47 (d, J ) 8.2, 1H), 7.38 (s, 1H), 7.35 (d, J ) 8.1, 1H), 7.28
(m, 1H), 7.22 (m, 2H), 7.13 (m, 1H), 6.93 (d, J ) 2.0, 1H), 3.17
(m, 2H), 3.08 (m, 2H); 13C NMR (125 MHz, CDCl3) δ 155.3,
141.3, 136.3, 128.3, 127.3, 124.1, 122.2, 122.0, 121.4, 120.3,
119.6, 119.3, 118.8, 115.9, 111.4, 111.1, 24.9, 24.4; MS (CI)
261 (61). 130 (100); HRMS (CI) calcd for C18H15NO 261.1153,
found 261.1153. Anal. Calcd for C18H15NO: C, 82.73; H, 5.79;
N, 5.36. Found: C, 82.73; H, 5.83; N, 5.32.
7.89 (m, 1H), 7.77 (d, J ) 8.3, 1H), 7.73 (d, J ) 8.2, 1H), 7.43
(m, 3H), 7.27 (m, 1H), 6.13 (d (br), J ) 6.7, 1H), 4.64 (s (br),
2H), 4.59 (s (br), 1H), 4.31 (m (br), 1H), 4.06 (dd, J ) 5.4, 11.1,
1H), 3.90 (s (br), 1H), 3.56 (m, 2H); 13C NMR (125 MHz,
DMSO-d6, 112 °C) δ 154.7, 140.7, 139.0, 125.9, 124.9, 124.0,
123.7 (br). 123.1 (br). 122.6, 121.5, 119.9, 119.5, 119.4, 114.7,
111.9 (br), 111.5, 110.9, 86.8 (br), 78.5, 77.7, 71.0 (br), 69.1,
68.3; MS (FAB): 389 (70), 257 (100); HRMS (FAB) calcd for
C23H19NO5 389.1263, found 389.1258.
12-Oxo-11-(â-D-t et r a -O-a cet ylglu cop yr a n osyl)in d olo-
[2,3-a ]ca r ba zole (17b). To 17a (0.15 g, 0.39 mmol) in pyridine
(2.0 mL) was added acetic anhydride (0.20 g, 2.0 mmol). The
reaction mixture was stirred at room temperature for 25 h and
then concentrated in vacuo. The residue was preadsorbed on
silica gel and purified by silica gel chromatography (30%
EtOAc/Hex) to provide 17b (0.19 g, 98%): [R]23 ) +144° (c
D
Ben zofu r ylin d olin e (16). To 15 (0.30 g, 1.15 mmol) was
added TFA (3.8 mL), and the reaction mixture was stirred for
15 min. After concentration in vacuo, the residue was taken
up into methylene chloride, and saturated NaHCO3 was added.
The solution was stirred vigorously for 10 min and then
extracted with CH2Cl2. The organic layers were combined,
washed with H2O and brine, and dried over MgSO4. Filtration
and evaporation of the solvent in vacuo provided 16 (0.28 g,
96%). An analytically pure sample was obtained by pread-
sorbing on silica gel and purifying by silica gel chromatography
(5% EtOAc/Hex): Rf ) 0.3 in 3% EtOAc/Hex; IR (KBr) 3431,
1.0, CHCl3); mp 228-230 °C (CHCl3); Rf ) 0.45 in 30% EtOAc/
Hex; IR (KBr) 3051, 2935, 2862, 1758 cm-1 1H NMR (400
;
MHz, DMSO-d6, 112 °C) δ 8.17 (d, J ) 7.7, 2H), 8.13 (d, J )
7.1, 1H), 7.96 (d, J ) 8.1, 1H), 7.86 (d, J ) 8.3, 1H), 7.81 (d, J
) 8.2, 1H), 7.51 (m, 3H), 7.30 (m, 1H), 6.53 (d, J ) 8.7, 1H),
6.02 (m (br), 1H), 5.62 (m, 2H), 4.29 (m, 1H), 4.01 (m, 1H),
2.10 (s, 3H), 1.97 (s, 3H), 1.33 (s, 3H); two carbon resonances
in the 110-155 ppm region and one carbon resonance in the
80-85 ppm range could not be resolved; 13C NMR (125 MHz,
DMSO-d6, 112 °C) δ 168.8, 168.7, 167.4, 154.8, 140.6, 139.1
(br), 126.2, 125.4, 123.9, 123.6 (br), 123.3 (br), 122.7, 121.7,
120.2, 120.1, 119.6, 114.9, 112.4, 110.8, 72.5, 70.1 (br), 67.9,
64.1, 19.7, 19.5, 18.6; MS (FAB) 515 (100), 257 (50); HRMS
(FAB) calcd for C29H25NO8 515.1580, found 515.1575.
1
3048, 2922 cm-1; H NMR (300 MHz, CDCl3) δ 7.42 (m, 2H),
7.21 (m, 3H), 7.02 (t, J ) 7.5, 1H), 6.76 (t, J ) 7.4, 1H), 6.63
(d, J ) 7.7, 1H), 4.89 (d, J ) 8.0, 1H), 4.40 (s (br), 1H), 3.83
(m, 1H), 2.65 (m, 2H), 2.32 (m, 1H), 2.14 (m, 1H); 13C NMR
(125 MHz, CDCl3) δ 154.7, 152.6, 150.1, 130.1, 127.8, 127.7,
123.9, 123.2, 122.3, 119.3, 119.2, 115.2, 111.0, 110.6, 55.6, 41.9,
24.0, 17.9; MS (CI) 261 (100); HRMS (CI) calcd for C18H15NO
261.1153, found 261.1143. Anal. Calcd for C18H15NO: C, 82.73;
H, 5.79; N, 5.36. Found: C, 82.56; H, 5.83; N, 5.26.
Ack n ow led gm en t. This work was supported by the
American Cancer Society RPG-97-159-01. Additional
support was provided by UC Cancer Research Coordi-
nating Committee, GlaxoWellcome, Eli Lilly, the Cam-
ille and Henry Dreyfus New Faculty award program,
and the Alfred P. Sloan Foundation.
12-Oxo-11-(â-D-xylop yr a n osyl)in d olo[2,3-a ]ca r b a zole
(17a ). To 16 (0.18 g, 0.69 mmol) in MeOH (5.0 mL) was added
D-xylose (0.31 g, 2.06 mmol). The reaction mixture was stirred
at reflux for 6.5 h and then concentrated in vacuo. The residue
was preadsorbed on silica gel, and the glycosylation products
were separated from the unused D-xylose by silica gel chro-
matography (2% MeOH/EtOAc) to provide a 1:1 mixture of
diastereomers (0.23 g, 85%). The mixture of diastereomers was
Su p p or tin g In for m a tion Ava ila ble: General experimen-
tal procedures and 1H NMR spectra for 1a , 1b, 9, 17a , and
17b. This material is available free of charge via the Internet
at http://pubs.acs.org.
J O990296V